Download Baryons at Low Densities: The Stellar Halos around Galaxies

Astronomical News
Report on the ESO Workshop
Baryons at Low Densities: The Stellar Halos around Galaxies
held at ESO Headquarters, Garching, Germany, 23–27 February 2015
Marina Rejkuba1
Magda Arnaboldi1
Elena Valenti1
1
ESO
The workshop (Stellar Halos 2015)
focused on theoretical and observational
studies of the faint surface brightness
envelopes surrounding luminous galaxies. This field of research has gained
much attention in recent years owing to
the availability of wide-field cameras,
deep and large-area surveys, as well as
advances in modelling galaxy formation
and evolution. The comparison of the
observational properties of stellar halos
with the predictions from simulations
provides important constraints on the
formation of luminous galaxies and their
assembly history. An overview of the
topics covered at the workshop is presented.
Stellar halos are ubiquitous in luminous
galaxies. On account of their low surface
brightness, detailed study of their physical properties has been difficult and, until
recently, confined largely to the Milky Way
(MW) and Andromeda (M31). Since the
advent of large cameras and surveys, both
from ground and space, our knowledge
of stellar halos has increased. For several
late- and early-type galaxies, astronomers
were able to measure the physical properties of their stellar halos out to hundreds
of kiloparsecs, or beyond, revealing very
low luminosity extended stellar structures, similar to the halos of the MW and
its closest neighbouring spiral M31.
Observations show that these halos
have complex morphologies with multiple
stellar components, complex kinematics
and substructures that indicate a history
of merger and accretion events. These
morphologies resemble the density maps
from cosmological simulations of galaxy
formation in a hierarchical Universe. Consequently a large amount of effort has
been invested in trying to understand how
we go from a qualitative resemblance to
quantitative measurements of the fractions
and frequencies of these substructures.
28
The Messenger 160 – June 2015
Figure 1. Participants of the Stellar Halos 2015 workshop in front of the ESO Headquarters building.
The Stellar Halos 2015 workshop brought
together around 90 observers and theorists (see Figure 1), who discussed results
from space- and ground-based surveys
of stellar halos in late- and early-type galaxies and their comparison with the
­predictions from cosmological simulations.
All the presentations were collected and
are now posted on the workshop web
page1, under Programme. A short, nonexhaustive, summary of the highlights is
presented here.
What is a galaxy stellar halo?
Guinevere Kauffmann and Annette
­Ferguson set the stage for the workshop
by addressing the question of what a
­galaxy stellar halo is from theoretical and
observational perspectives. Kauffmann
described the two main techniques for
the theoretical studies of stellar halos:
hybrid methods and full hydrodynamical
simulations. According to the hierarchical
cosmological model of Lambda Cold
Dark Matter (ΛCDM), the Milky Way and
other large galaxies of similar luminosity
should have accreted and subsequently
tidally destroyed ~ 200 low-mass galaxies
in the past ~ 12 Gyr. The simulated stellar
halo images produced by hybrid semianalytic plus N-body models, such as
those shown in Figure 2, from Bullock &
Johnston (2005), display the complex
structure of the halos as well as long-lived
low surface brightness streams. Simula-
tions can now trace the build-up of stellar
halos and the evolution of their properties.
For example, the projected phase-space
diagrams and plots of the line-of-sight
velocity vs. radial distance relative to the
host halo centre, can be used to identify
the recent satellite disruption events both
in simulations and in data (e.g., the talk
by Alessia Longobardi on the extended
substructure in the M87 halo — see also
Longobardi et al., 2015).
However, not all halo stars are accreted.
A fraction of the halo may be formed
in situ, either by heating the disc stars to
halo orbits, or by forming stars in the halo
from in-falling gas. The in-falling gas may
be stripped from sub-halos that merge
with the host galaxy, or enter through
direct infall. This dichotomy of “in situ” vs.
“accreted” halos has led to discussions
about the dual halo nature that was
emphasised by Annalisa Pillepich, who
showed results from a new set of simulations called Illustris. Andrew Cooper
showed the power of the particle tagging
method in making detailed predictions
for the statistical properties of stellar halos
that directly link CDM structure formation
to photometric and spectroscopic halo
observations.
Simulated images of stellar halos, like the
one shown in Figure 2, can be compared
with the observed surface brightness
maps around nearby galaxies. Observationally, halos can be studied by stacking
images of similarly bright objects, yielding
deep photometry of integrated light, and
by directly detecting individual stars and
Sagittarius dwarf provides ~ 20 % of all
the debris of the stellar halo in the Milky
Way, including multiple extended stellar
streams and of the order of 20 globular
clusters. The modelling of this stream
constrains the Galaxy potential and
its associated mass to ~ 5–8 × 1011 MA
within 200 kpc, in agreement with the
kinematics of halo stars and satellites in
the Local Group.
Figure 2. A simulated stellar halo surface brightness
image2 shown with a cut at 40 mag arcsec –2 (image
from Bullock et al., 2005).
other tracers. Yet, Annette Ferguson
emphasised that in order to measure
substructures in stellar halos, accurate
surface brightness photometry down to
µv ~ 30 mag arcsec –2, i.e., ~ 9–10 mag
(0.05 %) below the surface brightness
of the dark night sky, on angular scales
equivalent to ~ 100–200 kpc around
MW-sized galaxies, is required. This is
observationally very challenging! Since
different assembly histories for galaxies
may lead to complex, heterogeneous
and evolving structures, galaxy samples
made up of a range of morphological
types and environments are also necessary.
Stellar halos of the Milky Way and
Andromeda
The Milky Way and Andromeda are the
two nearest large galaxies whose halos
are being extensively studied. In his review
of the Milky Way stellar halo, Wyn Evans
quoted the fundamental observation from
Eggen, Lynden-Bell and Sandage (1962):
“The time required for stars in the halo
to exchange their energies and angular
momenta is very long compared with the
age of the Galaxy. Hence, knowledge of
their present energy and angular momenta
tells us something of the initial conditions
under which they formed.” Such a statement, which is more than half a century
old, illustrates very clearly the motivation
for studying the dynamics of streams in
galaxy halos. For example, the disrupted
Paula Jofre reported on measurement of
the ages of individual stars in the MW
halo, demonstrating that they are very
old, having been formed 10–12 Gyr ago.
The more metal-rich component tends
to have a higher fraction of younger stars.
There is also a bifurcation in the age–
metallicity relation seen in stellar clusters
and in field stars at [Fe/H] = –1.4 dex.
Giuseppe Bono presented a study of the
Galactic Halo metallicity gradient based
on a homogeneous set of RR Lyrae stars
with accurate distances. The RR Lyrae
stars are particularly suitable tracers of
the old Galactic Halo, showing a linear
shallow metallicity gradient. The younger
in situ component however is not traced
by RR Lyrae stars, which are all at least
10 Gyr old. Future spectroscopic surveys
of RR Lyrae stars will also enable comparison with simulations that predict different kinematical signatures for the in
situ vs. accreted components.
Two large surveys have investigated the
stellar halo of M31: the Pan-Andromeda
Archeological Survey (PAndAS) and
the Spectroscopic and Photometric
Landscape of Andromeda’s Stellar Halo
(SPLASH). The PAndAS results, presented by Rodrigo Ibata, are based on a
detailed photometric map extending
beyond the 150 kpc radius around M31
and covering also M33, using the
­Canada-France-Hawaii Telescope (CFHT).
The metal-poor halo of Andromeda is
found to be smoother than the metal-rich
part. Numerous dwarf galaxies and
­globular clusters as well as streams without clear progenitors, contribute to the
network of rich substructure in the halo.
Nicolas Martin, from the PAndAS survey,
is comparing the Andromeda field of
streams to Aquarius simulations, treating
the observational and model datasets
in a consistent way in order to reconstruct
the assembly history of the halo.
Karoline Gilbert presented the SPLASH
survey, which has traced the surface
brightness profile out to ~ 180 kpc, finding a similar density profile to the PAndAS
survey. The spectroscopic identification
of more than 1500 M31 halo stars in the
SPLASH survey has resulted in determination of the metallicity gradient out to
large radii. The stellar density and metallicity profiles indicate that the M31 halo
was largely built from one, or a few, relatively massive (> 109 MA) accretion events,
but at large radii many low-mass accretions have contributed to the recent buildup of the halo. The kinematics of the inner
halo stars is however consistent with their
origin from the disc — these stars were
likely kicked out in a collision with a system similar in size to the Large Magellanic
Cloud that collided with M31 ~ 760 Myr
ago. The MW and M31 metallicity and
α-element abundance gradients provide
datasets that can be used as constraints
to the simulations, such as those presented, for example, by Andreea Font.
Stellar halos in galaxies beyond the Local
Group
Laura Greggio reviewed the current
observational limitations and future opportunities for the studies of stellar halos.
She made the case that resolved stellar
population studies (single star photo­
metry) should reach well below the red
giant branch tip, hence arguing for deep
observations. The shapes of the substructures and surface density profiles
can then be used to assess the fraction
of accreted halo stars and recover the
individual galaxy accretion history. The
European Extremely Large Telescope
(E-ELT) and the James Webb Space Telescope (JWST) will open up the study
of resolved stellar halos to ~ 30 Mpc distance, such as red giant stars in the halos
of elliptical galaxies in Virgo. This ad­vance
will enable detailed studies of a statistically representative sample of large
­galaxies, and will connect the resolved
stellar population to the integrated light
studies.
Roelof de Jong and Antonella Monachesi
presented detailed physical properties of
individual galaxy halos beyond the Local
Group for the spiral galaxies in the ­Hubble
Space Telescope GHOSTS (Galaxy Halos
The Messenger 160 – June 2015
29
Astronomical News
Outer disks Substructure Thick disks Star
clusters) survey. The high angular resolution of the Hubble Space Telescope (HST)
allows the detection of single stars in
halos at larger distances and in denser
regions — hence investigating the transition regions between the disc and halo.
All MW-like galaxies observed by GHOSTS
have extended stellar halos. Their power
law profiles have a slope between – 2 < α
< – 4.5 over 10–70 kpc. The larger galaxies have fractionally larger envelopes and
the inner halos are compact, with Sersic
index n ~ 5, and are flat (c/a ~ 0.3).
Only a half dozen luminous early-type
galaxies are close enough to resolve individual stars in the halos and study their
properties. Marina Rejkuba summarised
the previous studies, most of which
concentrated on the nearest giant elliptical galaxy NGC 5128, known also as
­Centaurus A. The bulk of the stars in the
halo of NGC 5128 formed at redshift z > 2,
and these stars had very fast chemical
enrichment, already reaching super-solar
metallicity ~ 12 Gyr ago. This is the only
early-type galaxy in which metallicity
and stellar density gradients have been
mapped to 25 effective radii, or 140 kpc
distance. The halo appears to be elongated with a shallow metallicity gradient
and some field-to-field variation in stellar
density and metallicity, indicative of
­possible accretion events. This is not too
surprising, since this galaxy is a known
merger remnant.
On the other hand NGC 3379, a giant
elliptical in the Leo group, has little or no
indication of being disturbed or having
recently accreted stars in its halo. Beyond
~ 12 effective radii in this galaxy there
is a break in the metallicity distribution and
the metal-poor component, which is typically associated with halos, starts to dominate. Mark Peacock presented the results
from his HST study of NGC 3115, where
there is also quite a high mean metallicity,
ex­­tending to very large distances in the
halo: the metallicity decreases from –0.65
to –0.8 dex between 15–60 kpc (6–23
effective radii). However, a distinct metalpoor population is identified in the two
outer fields and its profile is found to
trace the density profile of the metal-poor
globular clusters. The metal-poor halo
population mass fraction is ~ 14 % of the
total halo mass in NGC 3115.
30
The Messenger 160 – June 2015
Rejkuba M. et al., Report on the ESO Workshop “Baryons at Low Densities”
3$QG$600DS
0F&RQQDFKLHHWDO
1
1
1
(
0
NSF NSF
NSF
NSF
*LDQW6RXWKHUQ
6WUHDP
0G6SKV
0+DOR)LHOGV
0
'ZDUI*DOD[\)LHOGV
Figure 3. Map of the stellar counts in the Andromeda
galaxy M31 from the PAndAS survey (McConnachie
et al., 2009) with overlayed spectroscopic observation target fields from the SPLASH survey 3.
Never­theless tremendous progress has
been achieved and there are ever-larger
samples of galaxies where faint halos and
streams are detected.
More distant early-type galaxies, in the
Virgo and Fornax clusters, have too high
a density of stars for accurate resolved
stellar population studies, hence deep
surface brightness observations are
favoured for the study of their faint envelopes. Chris Mihos has traced the envelope of M49 at the centre of the Virgo subcluster B out to ~ 150 kpc. A steep bluer
colour gradient has been measured in
the outer regions indicating a very high
fraction of metal-poor stars in its outskirts.
Enrichetta Iodice and Marilena Spavone
examined the extended envelopes of
individual Virgo and Fornax cluster galaxies from VLT Survey Telescope and
OmegaCAM guaranteed time observation
programmes. The CFHT was used by
Pierre-Alain Duc to systematically survey
the extended regions around galaxies
whose internal kinematics and stellar populations were observed by the A
­ TLAS3D
project. Several instrumental effects may
limit the accuracy of these measurements:
flat-fielding, scattered light, corrections
due to the wings of the point spread
function and halos due to internal reflections in the camera, as well as Galactic
cirrus, must be taken into account.
Particularly impressive accretion streams
were presented by David Martinez-­
Delgado, based on imaging with small
(0.1–0.5 metre) robotic telescopes. These
observations reveal structures around
galaxies with a striking resemblance to
those shown by cosmological simulations
of structure formation. Richard D’Souza
presented the statistical properties of
stellar halos. By carefully stacking a very
large number of galaxies, he found that
the stellar halos of late-type galaxies tend
to be spherical, while those of early-type
galaxies tend to be elliptical. The ellip­
ticity increases with stellar mass, and
also the fraction of light in the outskirts is
higher in the more massive galaxies.
The surface density profiles agree with
the results from simulations using particletagging methods.
Magda Arnaboldi, Pat Durrell and Jean
Brodie presented the use of discrete
­tracers, like planetary nebulae and glob­
ular clusters, to study the physical
­properties of the extended halos. Durrell
and Brodie illustrated how red and blue
globular clusters trace different com­
ponents in the halos. The red globular
Astronomical News
cluster populations have number density
profiles similar to that of the stars, while
the blue globular clusters are spatially
more extended and seem to trace the
metal-poor halo component. Arnaboldi
showed the overlap of the M87 halo and
the intracluster light in the radial range
between 20 to 150 kpc from the centre of
M87, and how the planetary nebulae
luminosity function is used to identify the
star-forming (metal-poor) stellar population
vs. the old and metal-rich one. Michael
Hilker showed that there are different subpopulations of globular clusters in the
distant central galaxy in the Hydra cluster
that trace different assembly epochs.
The origin of ultra-compact dwarfs around
NGC 1399, the central galaxy of the
­Fornax cluster was presented by Karina
Voggel. Ortwin Gerhard, Nicola Napolitano
and Eric Emsellem described the kinematics of stars, their angular momentum
and orbital distribution in galaxies from
the centre out to ten effective radii.
While the stellar mass is a strongly nonlinear function of the galaxy halo mass,
due to the large contribution of dark
­ atter in the least and the most massive
m
galaxies, the total mass of all globular
clusters correlates linearly with the mass
of the halo, as shown by Bill Harris. This
is explained if globular cluster formation
is largely immune to the feedback that
affects star formation in the field, or if
the formation of globular clusters is completed before the feedback starts.
Prospects
During the final discussion, the obvious
energy and enthusiasm testified that
this field of research had gained sub­
stantial momentum! The observational
samples are growing and we are starting
to make quantitative comparison of the
properties of halos inferred by different
observational techniques with the simu­
lations. The importance of consistent
comparisons was emphasised yet again:
there are different ways to define a stellar halo — including or excluding the
streams and accreted material — which
then lead to different halo properties.
Hence the halo properties need to be
studied in a homogeneous and unbiased
way, by applying a consistent treatment
to observations and theory/simulations.
Lively discussions during the workshop,
after every talk, during coffee breaks,
and the vibrant atmosphere during the
final gathering around Bier und Bretz’n,
resulted in new ideas, new collaborations
and projects for future advances in the
field.
References
Bullock, J. S. & Johnston, K. V. 2005, ApJ, 635, 931
Eggen, O. J., Lynden-Bell, D. & Sandage, A. R. 1962,
ApJ, 136, 748
Longobardi, A. et al. 2015, A&A, in press,
arXiv:1504.04369
McConnachie, A. W. et al. 2009, Nature, 461, 66
Links
1
orkshop programme: http://www.eso.org/sci/
W
meetings/2015/StellarHalos2015/program.html
Image of simulated stellar halo: http://www.physics.
uci.edu/~bullock/StellarHalo/halo.html
3
A nnotated image of M31: https://blogs.stsci.edu/
universe/2014/07/21/what-built-andromedas-stellarhalo/
2
Report on the ESO Workshop
Dissecting Galaxies Near and Far
held at ESO Vitacura, Santiago, Chile, 23–27 March 2015
Catherine Vlahakis1,2
1
J oint ALMA Observatory, Santiago,
Chile
2
ESO
The workshop explored high-resolution
perspectives for the understanding
of star formation and the interstellar
medium in the era of ALMA. Angular
and spectral resolutions are fundamental limitations for our understanding of
the properties of galaxies in the nearby
and distant Universe. A new generation
of instruments and powerful observa­
tories, as well as planned future facili-
ties, is pushing out the boundaries of
what was previously possible over the
entire electromagnetic spectrum. Held
in the approach to ALMA Early Science
Cycle 3, the workshop centred on discussing the scientific results and
upcoming opportunities afforded by the
new capabilities available in the era of
ALMA. Some of the science highlights
of the meeting are presented.
The advent of state-of-the-art facilities
with high angular resolution and/or high
spectral resolution capabilities, coupled
with high sensitivity, such as the Atacama
Large Millimeter/submillimeter Array
(ALMA), allows the interstellar medium
(ISM) and star formation properties of galaxies at intermediate redshift to be probed
in unprecedented detail, along with their
counterparts in the local Universe.
The aim of the workshop was to bring
together astronomers whose interests
further our understanding of star formation and ISM processes at high spatial
and spectral resolution, both in nearby
and distant galaxies. The theme of
­“dissecting galaxies” was intended not
only to encompass what we can learn
by going to higher spatial and spectral
resolution, but also, in the true sense of
the word “dissect”, what we can learn
about galaxies when we methodically
break them down into their constituent
The Messenger 160 – June 2015
31